Modified zwolle score with delta-creatinine: enhancing the safety of early discharge after STEMI

Abstract

Introduction The Zwolle score (ZS) is recommended to identify low-risk patients eligible for early discharge after acute ST-segment elevation myocardial infarction (STEMI), but as only one-third of STEMIs have a low ZS, the discharge is often postponed. Creatinine variation (Δ-Cr) also provide prognostic information after STEMI. Purpose The authors intend to study the “modified Zwolle Score” (MZS) model, which encompasses Δ-Cr as a variable that may enhance the discriminative power of the standard ZS. The outcome is 30-day mortality, time range that starts right after the ACS. Methods This is a retrospective study with data from a national multicentre registry. We have included 3.296 patients with STEMI. Zwolle score was calculated for each patient. It is defined as shown in figure 1. Δ-Cr was defined as maximum serum creatinine minus admission serum creatinine. A Δ-Cr≥0.3 was assigned 2 points in the Modified Zwolle Score, after interpretation of odds ratio via multivariate analysis. For prediction quality assessment, we have performed ROC curve analysis with both scoring systems versus 30-day mortality. Regarding survival analysis, we have performed Kaplan-Meier curves with Log-rank analysis. We have also registered complications during hospital stay. Results The sample mean age is 63±14, and it is composed by 76.8% of males. The majority of patients presented Killip Class I (87.3%). The STEMI was anterior in 49.7% of patients and inferior in 49.8% of patients. The mean admission time was 5 days. Intrahospital mortality was 3% and 30-day mortality was 4%. The mean ZS was 3.1±2.8 points, the mean MZS was 3±2.1 points and the mean Δ-Cr was 0.2±0.6mg/dL. The ROC curve analysis between ZS and early mortality revealed a c-statistic of 0.810 (CI 0.796–0.823), whereas the ROC curve between MZS and early mortality revealed a c-statistic of 0.853 (95% CI: 0.841–0.865). The ROC curves comparison showed superiority of the MZS c-statistic, with a difference between AUC of 0.043 (p<0.001, 95% CI: 0.024–0.063). Regarding low-risk patients, 30-day mortality was 3.3% using ZS (0–2 points) and 2.4% using modified ZS (0–2 points). Fifty patients (1.5%) died between 3rd and 10th day after ACS: original ZS low-risk criteria registered 0.09% and modified ZS low-risk criteria 0.06% fatalities. Kappa coefficient for intergroup concordance was good (0.73). Conclusion We conclude that by adding Δ-Cr to the standard ZS, a renal function parameter that was lacking in the ZS, its predicting capacity regarding early mortality in patients admitted with STEMI was increased. Comparing both scores, low-risk patients defined by MZS registered less complications, 3–10 day mortality and 30-day mortality than low-risk patients defined by the original ZS. This fact may lead to better distinction of patients who will benefit from early discharge. Zwolle Score, ROC curves and survival Funding Acknowledgement Type of funding source: None